# Load with dill
import dill
with open('schuheinlagen_orakel.dill', 'rb') as f:
    ki_modell = dill.load(f)

import pandas as pd 

bauteil = pd.DataFrame({
    'Zellenform': [1],
    'Zellengroesse': [3.0],
    'Fuellgrad': [0.3]
})

bauteil.head()

	Zellenform	Zellengroesse	Fuellgrad
0	1	3.0	0.3

import numpy as np

N = 100
variation_zellenform = np.ones(N)

np.random.seed(42) 

variation_zellengroesse = 3.0 + np.random.normal(0, 0.5, N)
variation_fuellgrad = 0.3 + np.random.normal(0, 0.05, N)

eingabedaten = pd.DataFrame({
    'Zellenform': variation_zellenform,
    'Zellengroesse': variation_zellengroesse,
    'Fuellgrad': variation_fuellgrad
}) 

ausgabedaten = ki_modell.predict(eingabedaten)

abstaende = ((eingabedaten['Zellengroesse'] - 3.0)**2 + (eingabedaten['Fuellgrad'] - 0.3)**2)**0.5
gewichte = -2/3 * abstaende + 1

from sklearn.linear_model import LinearRegression

modell = LinearRegression()
modell.fit(eingabedaten, ausgabedaten, sample_weight=gewichte)
score = modell.score(eingabedaten, ausgabedaten)
print(score)

print(modell.coef_)

0.8904881638530242
[  0.          -2.45507152 635.33023841]

eingabedaten['Fuellgrad'].to_list()

[0.22923146289747925,
 0.278967733861732,
 0.2828642741736615,
 0.259886136538919,
 0.2919357144166995,
 0.3202025428407269,
 0.3943092950605265,
 0.3087288906415919,
 0.3128775195361382,
 0.29627770421169164,
 0.2040614392350479,
 0.29867430622753915,
 0.30301151049705133,
 0.4231621056242643,
 0.29038195176094383,
 0.3150773671166806,
 0.2982644115147378,
 0.24156609811902338,
 0.35714114072575104,
 0.3375966516343387,
 0.33955159735215235,
 0.25453062726026304,
 0.37013971554680497,
 0.22990744686038594,
 0.3293428546900135,
 0.4095227812904989,
 0.25047318374346556,
 0.2716851135198614,
 0.304982568254382,
 0.27482621729419004,
 0.22246682844669335,
 0.30342814874030133,
 0.24688481431369474,
 0.32367962153175905,
 0.2540287882883098,
 0.377496720250877,
 0.26083733538318815,
 0.28389692418971624,
 0.34067586086848345,
 0.23845678417830224,
 0.31137299673020646,
 0.36535713771412137,
 0.2196258382719386,
 0.3092316929266152,
 0.3129941397124212,
 0.3390911435888655,
 0.2381524644560959,
 0.23397716934578616,
 0.3260970782808449,
 0.3148492336616593,
 0.31252464251729384,
 0.31732241047484877,
 0.26599876392107547,
 0.31161268485805016,
 0.31465362366493405,
 0.2642824290986816,
 0.3932887255572378,
 0.32369164604558937,
 0.24043482513986755,
 0.3328276804316915,
 0.25126591648863394,
 0.3393542301871226,
 0.3579297789503702,
 0.25896588408241444,
 0.3481688064622161,
 0.3206390463468249,
 0.3411030079997245,
 0.39483964913269737,
 0.28773059419985647,
 0.26231319178212553,
 0.25552427851872384,
 0.2592094857517281,
 0.29614491452929476,
 0.31705759874083217,
 0.31383453996650096,
 0.3413591624518012,
 0.30065009459389536,
 0.37267670385786583,
 0.2867671583381022,
 0.43600845832948093,
 0.3312833673882503,
 0.25714212217918586,
 0.24645537509694437,
 0.3241236207621593,
 0.28882686073370745,
 0.3357000247046046,
 0.32366188122867723,
 0.29635855436715636,
 0.25766031409657975,
 0.22425763876570676,
 0.27767425239664895,
 0.34281993971617364,
 0.3107046872065102,
 0.2377130610644006,
 0.3086590462925591,
 0.3192658689864418,
 0.25580712818994333,
 0.3076862552972764,
 0.3029104359223,
 0.24285148510846882]

import plotly.express as px
import plotly.graph_objects as go
from ipywidgets import interact, FloatSlider
import numpy as np


# Scatter plot
fig = px.scatter(eingabedaten,
    x='Fuellgrad',
    y=ausgabedaten,
    title='Regressionsgerade für maximale Kraft abhängig vom Füllgrad',
    labels={'y': 'maximale Kraft[N]'}
)

# Interactive function for updating the regression line
def update_regression_line(weight=1.0, intercept=0.0):
    fig.data = []  # Clear existing data to avoid overlapping lines
    fig.add_trace(go.Scatter(x=eingabedaten['Fuellgrad'], y=ausgabedaten, mode='markers', name='Daten'))

    # Calculate regression line based on slider values
    regression_line = weight * eingabedaten['Fuellgrad'] + intercept
    fig.add_trace(go.Scatter(
        x=eingabedaten['Fuellgrad'],
        y=regression_line,
        mode='lines',
        name='Regressionsgerade',
        line=dict(color='red')
    ))
    
    fig.update_layout(xaxis_range=[0.15, 0.45], yaxis_range=[0,250])
    fig.show()

# Sliders for controlling weight and intercept
interact(
    update_regression_line,
    weight=FloatSlider(value=1.0, min=500, max=600, step=5, description='Steigung'),
    intercept=FloatSlider(value=0.0, min=-120, max=-100, step=1, description='y-Achsenabschnitt')
)

<function __main__.update_regression_line(weight=1.0, intercept=0.0)>